VOL.
5,
NO.
8,
AUGUST
1966
Side-ChainCotton Effects of Ribonuclease* Robert T. Simpsont and Bert L. Vallee
ABSTRACT: Chemical modifications, solvent, and pH variation suggest that the side-chain Cotton effect of native ribonuclease observed at neutral pH is associated with one or more of the buried tyrosyl residues. At pH 11, the ionization of surface tyrosyl residuds) generates a second Cotton effect not apparent when the residue(s) is protonated. Hence, at such alkaline pH, both these two Cotton effects are present but unresolved. Acetylation of ribonuclease with N-acetylimidazole in aqueous solution modifies three tyrosyl residues without significant effect upon either enzymatic activity or the optical rotatory dispersion of the protein. In contrast, acetylation in urea, followed by its removal, modifies all six tyrosyl residues and abolishes both the activity and the side-chain Cotton effect. The optical rotatory dispersion of this modified protein and the kinetics of deacylation by hydroxylamine suggest that acetylation of the buried tyrosyl residues
S
tudies of a number of proteins by optical rotatory dispersion (Simmons and Blout, 1960; Ulmer et a/., 1961; Beychok, 1965; Cathou et al., 1965; Coleman, 1965; Glazer and Simmons, 1965a,b; Myers and Edsall, 1965; Ulmer, 1965; Ulmer and Vallee, 1965; Urry, 1965) and amino acid polymers (Fasman et al., 1964, 1965) have recently demonstrated the widespread occurrence of side-chain Cotton effects between 250 and 300 mw, Apart from their intrinsic significance, the origin of such Cotton effects is important for the interpretation of the remainder of the rotatory dispersion spectrum. Further, since these Cotton effects are conformation dependent, they constitute a sensitive means of evaluating the role of aromatic amino acids in the formation of the tertiary structure of proteins. The possibility that tyrosyl residues of proteins might be optically active has provided a new approach for the investigation of the structural and functional roles of tyrosine in enzymes. Ribonuclease has previously served importantly to elucidate the structural and functional role of tyrosyl residues. The evidence supports the view that three tyrosyl residues of ribonuclease are “free” and exposed
* From the Biophysics Research Laboratory, Department of Biological Chemistry, Harvard Medical School, and the Division of Medical Biology, Peter Bent Brigham Hospital, Boston, Massachusetts. Receiced May 6, 1966. This work was supported by Grant-in-Aid HE-07297 from the National Institutes of Health of the Department of Health, Education and Welfare. t Postdoctoral Fellow of the National Institutes of Health.
prevents the refolding of the enzyme on removal of urea. Deacylation restores both activity and the rotatory dispersion pattern to resemble those of native ribonuclease. Solvent variation further demonstrates the interdependence of the environment of the buried tyrosyl residues and the occurrence of the side-chain Cotton effect. At alkaline pH, the apparent midpoint of the side-chain Cotton effect shifts to longer wavelengths. Optical rotatory dispersion and circular dichroism of the native protein at near-neutral pH compared with such determinations at alkaline pH indicate that the ionization of surface tyrosyl residue(s) generates a new Cotton effect, in addition to that associated with the buried tyrosyl residues. The consequences of chemical modification of buried tyrosyl residues are discussed in terms of their role in determining secondary and tertiary structure and maintainance of the conformation of ribonuclease.
to the solvent, while three are ‘‘buriedl’l within the interior of the folded molecule (Scheraga and Rupley, 1962). The three buried tyrosyl residues have been thought to function in maintenance of the tertiary structure of ribonuclease through hydrogen-bond formation with aspartyl carboxyl groups (Scheraga, 1957, 1960), or, alternatively, through hydrophobic interactions (Tanford et al., 1955). Since ribonuclease does not contain tryptophan and the conditions for the selective visualization of its free and buried tyrosyl residues have been established, studies of optically active aromatic absorption bands of this enzyme might assist in delineating the role of free and buried tyrosyl residues in generating side-chain Cotton effects. In this regard we have previously reported on the use of N-acetylimidazole both for the identification of free tyrosyl residues in proteins (Wacker et al., 1964; Riordan et al., 1965a) and for the study of their role in enzymatic function (Simpson et al., 1963; Riordan et al., 1965b). The data obtained in this manner were consistent with those obtained by other experimental means. The present study reports the effects of acetylation with this agent on the optical rotatory dispersion and enzymatic activity of ribonuclease. The effect of various nonaqueous solvents and pH upon the optical rotatory dispersion has also been examined. Finally, the circu-
1 The terminology here employed has been discussed previously (Riordan et al., 1965a)
SIDE-CHAIN
COTTON
EFFECTS OF
2531
RIBONUCLEASE
BIOCHEMISTRY
lar dichroism of the protein at neutral and alkaline pH has been evaluated. The data suggest that the aromatic side-chain Cotton effect of ribonuclease observed at 278 mp and at neutral pH should be assigned to the buried tyrosyl residues which stabilize the tertiary structure of the enzyme. In addition, however, a free tyrosyl residue generates a Cotton effect when ionized. The Cotton effect observed at pH 11.5, where only the free residues are ionized (Shugar, 1952), seems to be the sum of both of these. The superimposition of these two Cotton effects apparently produces the bathochromic shift in the midpoint of the side-chain Cotton effect at pH 11.5 previously observed (Glazer and Simmons, 1965b). Experimental Section Lyophilized ribonuclease A, isolated by chromatography and free of phosphate, was obtained from the Worthington Biochemical Corp., Freehold, N. J. N-Acetylimidazole (Aceto Chemical Corp.) was stored over phosphorous pentoxide in cacuo after recrystallization from dry benzene or isopropenyl acetate. Urea was recrystallized from 95 % ethanol and air dried, and solutions were prepared immediately prior to use. All other chemicals were reagent grade. Ribonuclease concentrations were determined spectrophotometrically based upon a molar absorptivity of E~,,.~ 9.8 X l o 3 M-' cm-' (Sela and Anfinsen, 1957), and a molecular weight of 13,683 (Hirs et al., 1956). Concentrations of the acetylated enzymes were determined following deacetylation with 1 M hydroxylamine (aide infra).
2532
Acetylation of ribonuclease was carried out by addition of an 180-fold molar excess of dry N-acetylimidazole to 4.5 mg of ribonuclease/ml either of 0.02 M sodium Veronal or 0.02 M sodium Veronal plus 8 M urea, pH 7.5,23". The reaction was allowed to continue for 45 min. Solutions were then dialyzed in pretreated Visking cellulose dialysis tubing (Callanan et al., 1957) against large excess of 0.1 M phosphate buffer, pH 6.1 or 7.0,4", for three changes of at least 8 hr each. The degree of modification of tyrosyl residues was determined by measurements of spectra based on Ae2,s 1160/mole of tyrosyl residues acetylated (Simpson et al., 1963). Modification of amino groups was assayed using an automated ninhydrin method. Samples of 0.1 ml, 4 mg/ml, were diluted into 0.9 ml of 0.2 M citrate buffer at pH 5.0, sampled at 0.42 mljmin and mixed with ninhydrin reagent (Lenard et al., 1965) at a flow rate of 1.2 ml/min. Color was developed at 100" for 15 min and quantitated photometrically at 550 mp in an Auto-Analyzer system (Technicon Instrument Corp., Chauncey, N. Y . ) , using phenylalanine as a standard. Deacetylation of the modified proteins was performed either with 1 M hydroxylamine at 25" in pH 7.5, 0.1 M phosphate buffer for 10 min, or alternatively, with 0.1 M hydroxylamine for 1 hr in the same buffer at 25". Labile acetyl groups were determined as previously described (Simpson et al., 1963) except that 0.1
R O B E R T T.
SIMPSON
A N D
B E R T L.
VALLEE
phosphate buffer, pH 7.5, was employed. Ribonuclease activity was determined with soluble yeast ribonucleic acid as a substrate as described by Anfinsen et al. (1954). Assays were performed at 25" with 5-10 pg of ribonuclease/ml of 0.1 M sodium acetate buffer, pH 5 . pH was determined with a Radiometer pH meter using a general-purpose glass electrode. Absorbance measurements at discrete wavelengths were determined with a Zeiss PMQII spectrophotometer, while continuous absorption spectra were obtained with either a Cary Model 11 or Model 15 automatic recording spectrophotometer. Optical rotatory dispersion measurements were performed in a Cary Model 60 automatic recording spectropolarimeter at 25 ". Generally, measurements were performed with a 5-mm cell and protein concentration of about 0.4% from 350 to 250 mp, with a 0.5-mm cell and the same protein concentration from 220 to 280 mp, and with a 0.5-mm cell and a protein concentration of 0.08% from 200 to 225 mp. The slit width of the instrument was programmed to yield constant light intensities at all wavelengths. In areas of high absorbance, absolute values for specific rotation were confirmed at two or more protein concentrations or path lengths, eliminating the possibility of spurious Cotton effects (Urnes and Doty, 1961). The data are expressed as specific rotation in degrees and are not corrected for the refractive index of the solvents employed. Circular dichroism measurements were performed with a Durrum-Jasco spectropolarimeter with a circular dichroism attachment. Conditions employed were identical with those used for the rotatory dispersion measurements. The results are expressed as the observed AAL-R for a 0.4% solution of ribonuclease in a 5-mm cell. M
Results The ultraviolet absorption spectra of native ribonuclease and ribonuclease acetylated both in the presence and the absence of 8 M urea are shown in Figure 1. The hypo- and hypsochromic shift in the absorption of tyrosyl residues, previously demonstrated, consequent to acetylation of the phenolic hydroxyl group is apparent (Schlogl et al., 1953; Simpson et al., 1963). Upon acetylation in aqueous solution, the absorption maximum is shifted from 277.5 for the native enzyme to 279 mp, and the absorptivity at 278 mK is markedly decreased. The residual spectral properties of this modified protein in the region of 280 mp largely reflect the absorption of the remaining unmodified tyrosyl residues. Based on the change of molar absorptivity at 278 mp, Aens 1160, for the conversion of N-acetyltyrosine to N,O-diacetyltyrosine, three tyrosyl residues are modified in this instance, in accord with previous findings (Riordan et al., 1965a). This derivative will be referred to as Ac3RNAase. Ribonuclease (RNAase) acetylated in the presence of 8 M urea absorbs maximally at approximately 262 mp (Figure 1). The magnitude of the decrease in absorptivity at 278 mp demon-
VUL.
5,
TABLE I:
NO.
8,
A U G U S I
I966
Characteristics of Modified Ribonucleases.. Act. Sample
(% control)
Native RNAase AcaRNAase AcsRNAase Deacylatedh Ac6RNAase
100 85
H ydroxamates (mole/mole)
0-Acetyltyrosyls (mole/mole)
Amino Groups (% control)
0 3 0 5 9 0
0 3 0 6 0 0
100 21 22 22
